As an EU citizen and AI practitioner, I couldn't help but get excited about everything implemented around the the 2025 AI Action Summit this week:

Public-Private Investment Acceleration (2025-2030)

The European Commission committed €50 billion in public funds to complement a €150 billion private sector pledge in InvestAI, making it the largest AI-focused public-private partnership globally.

This €200 billion EU-wide envelope includes:

• AI Industrialization: 65% allocation for manufacturing/healthcare integration, with €7B dedicated to semiconductor R&D via the Chips Joint Undertaking.
• Compute Infrastructure: France's €109 billion package (€30-50B from UAE for data centers, €20B from Brookfield, €3B from Mistral AI) targeting 1GW nuclear-powered capacity by 2026.
• SME Access: €20B venture debt facility through European Investment Bank, launching Q3 2025 with 2% interest subsidies.

The Current AI Initative (Operational Q4 2025)

A €400 million initial investment pooling public funds (60%), philanthropic capital (25%), and corporate contributions (15%) to put human and public interest back front and centre in advancements in AI:

• Create 78 industry-specific datasets for healthcare diagnostics (Owkin), precision agriculture (Ynsect), and renewable energy optimization.
• Develop open-source compliance tools for EU AI Act requirements, including real-time risk assessment modules.
• Establish AI impact measurement framework (Social Return on AI Index) for ESG reporting.

For more information check out the Current AI website and press release.

The AI Transition Fund (2025-2028)

€20B allocation targeting:

• Up-skilling 2M workers in prompt engineering/maintenance via the Digital Europe Programme.
• Creating 45 regional AI innovation hubs (€300M each) connecting vocational schools with corporate partners.
• Launching EU-wide certification system for AI specialists (Q1 2026).

The Coalition for Environmentally Sustainable AI (CESA)

A 91-member initiative to make sure EU AI energy needs are met sustainably:

• 2026: Real-time energy tracking for data centers > 50MW (Mistral’s French facilities first).
• 2027: Carbon budgeting for models >10B parameters (max 550tCO2eq annual allowance).
• 2028: 60% recycled rare earth minerals in AI hardware through CSRD amendments.

The Mistral-Helsing Defense AI Consortium

A €3B joint venture developing vision-language-action models for:

• Autonomous drone swarms (2026 field testing)
• Predictive maintenance of Leclerc tanks (2027 rollout)
• NATO-compliant cyber threat prediction (2028 operational)

Regulatory Sandbox Expansion

France’s “Notre Dame Strategy” creating fast-track pathways:

• 60-day approval for AI clinical trials (vs 180-day EU average).
• 50% R&D tax credit through 2030 for sustainability-focused AI.
• Data localization exemptions for GDPR-compliant third countries (US/Canada/Japan).

The Bigger Picture

I'm thrilled to see the EU finally taking AI seriously, and - to be honest - I'm a bit taken by surprise by the extent of the commitments at and around the AI Action Summit.

Positively surprised for sure, because in my mind Europe entering the race for AI leadership is great news for everyone - not just for those working in AI:

• The current two-horse race for AI between the US and China makes AI seem like a zero-sum game, distracting from real benefits and impact of this transformative technology.
• The European approach to human-centric value driven AI gives us more data points and perspectives on how to integrate AI into our lives and businesses.
• The proposed investments will lead to tons of jobs and opportunities in and around European tech, benefiting the wider EU economy.

This isn't just about competition. It's about shaping AI for good, globally.

---

No-Code AI Apps: It's Become Easier Than Ever To Build Your Own Apps

The best applications of LLMs (Large Language Models) empower non-technical folks to change their lives in ways that would have never been possible before the recent advancements in AI.

One of the tools in this bucket is Lovable AI, a platform that makes building sophisticated AI-powered apps as easy as assembling IKEA furniture.

Let's look at a couple of examples to see what Lovable does.

Kyler’s AI Hiring Platform: Building The Dream

Meet Kyler, a recruiter with an impressive résumé. He spent nearly a decade hiring top talent for companies like Google, Unity, and Rivian. But deep down, Kyler had always dreamed of building something himself.

That all changed when Kyler discovered Lovable AI. Using its no-code platform, he built an AI-powered hiring platform that automates key recruitment tasks—resume analysis, job matching, outreach, and even blockchain-based credential verification.

What would have taken months of engineering work (and hundreds of thousands of dollars) was completed in just days.

Here’s what Kyler’s platform does:

• AI Resume Builder: Generates optimized resumes from LinkedIn profiles.
• Job Matching: Compares resumes to job descriptions and identifies missing skills.
• Automated Outreach: Drafts personalized messages for hiring managers at scale.
• Blockchain Credentials: Verifies candidate records for fraud-proof hiring.
• Smart Sourcing: Finds top talent from platforms like LinkedIn and GitHub based on employer needs.

The results? Kyler reduced hiring friction by 40% and launched his platform with features that would’ve taken six months to build traditionally—completed in just four days! His innovation is already turning heads, with major enterprises exploring blockchain-based hiring solutions and 1,300 token holders backing his cryptocurrency-powered marketplace.

Harsh’s HealthSync: Revolutionizing Wellness in Just Six Hours

Or what about Harsh Patel—a software developer, AI enthusiast, and the creator of HealthSync, which he built in just six hours using Lovable AI. Yes, you read that right—six hours.

Picture this: You’ve got data scattered across Fitbit, Apple Health, and other apps, but no way to tie it all together. That’s where HealthSync comes in. Harsh’s app aggregates all your health data into one seamless portal and uses AI to provide personalized insights.

HealthSync

• Syncs with health apps like Fitbit and Apple Health to pull in all your activity and nutrition data.
• Analyzes your metrics to give actionable advice—like how many calories you need to consume or steps you should aim for.
• Breaks down meals by calories, carbs, proteins, and more.
• Recommends vitamins, routines, and fitness goals based on your specific data.

The magic? It’s not just about tracking—it’s about actionable advice. The app doesn’t just tell you what you’ve done; it tells you what you should do next to improve your health.

How Did Harsh Build It So Fast?

1. Prompt Engineering: Harsh crafted detailed prompts specifying the app’s structure—progress screens, AI insights, meal analysis, and more.
2. Lovable’s Code Generation: Lovable generated the foundational code for the app based on his prompts.
3. Database Integration: He used Supabase for the database portion, which integrated seamlessly with Lovable’s platform.
4. Fine-Tuning: After generating the initial code, Harsh refined it until everything clicked perfectly.

According to Harsh: “Software that would have taken months to develop was ready in hours thanks to Lovable.”

Why No Code AI Is Your New Secret Weapon

These stories are proof that today anyone can build an app that makes a difference in their lives - democratising not just AI, but technology access without having to spend years learning the craft of software development.

Here’s why apps like Lovable AI are changing the game:

• Ridiculously Easy: If you can drag-and-drop, you can build an app.
• Time-Saving Magic: Create powerful tools in hours instead of months.
• Budget-Friendly: Say goodbye to expensive developers.
• Customizable: Build something tailored perfectly to your needs.

Whether you’re solving hiring headaches like Kyler or transforming healthcare like Harsh, this platform is a great way to bring your ideas to life.

---

This week in AI

• The AI Action Summit Paris 2025 saw a gathering of who's who in European AI along with pledges for a massive ramp-up in investments and public-private AI partnerships across Europe with the goal of positioning the EU as a leader in AI in the next 5 years.
• The European InvestAI Initiative was announced at the summit by the European Commission: a €200B investment initiative, including €20B for AI gigafactories, looking to position Europe as a global leader in collaborative AI model development.
• OpenAI has announced plans to release GPT-5 in 2025, unifying its GPT and o-series models into a single system. The model will feature advanced reasoning, voice interaction, a visual canvas for images and text, integrated search, and research tools. Free users will access GPT-5 at standard intelligence, while Plus and Pro subscribers get enhanced capabilities. This streamlined approach eliminates the "model picker," creating a seamless AI experience. Before GPT-5, OpenAI will launch GPT-4.5 (Orion) as the final non-chain-of-thought model in the coming weeks.
• Latent Labs, founded by a former DeepMind scientist, raised £40.4M to advance generative AI models for designing therapeutic molecules like antibodies and enzymes. This innovation aims to accelerate drug discovery and improve success rates.